For the higher demand in multimedia quality and function of entertainment and consumer electronic devices, the tendency of the multimedia technology is toward high speed and efficient computation. The related hardware and software resources become essential to the real-time audio and video signal processing development. Hence, the design and manufacture of multimedia product tend to low cost, high mobility, and so on.
In general, the multimedia signals consume more resources for storage requirement, and a number of different image compression standards such as DV (digital video) standard specified in the ISO/IEC 61834, the “Blue Book”, and the SMPTE 314 specification, for encoding video content have been used. The device employs the DV format classifying pixels into blocks during encode stage in some mathematical way such as a DCT (discrete cosine transformation) technique for DCT block obtainment, compressed data representation in DCT coefficient, and loss reduction of image retrieval in quantization, whereas a decoder of the device performs reverse procedure for approximate DCT coefficient retrieval during decode stage.
Besides, the video data stream with the DV format includes several digital video frames have thirty segments with five macro-blocks respectively. Moreover, each macro-block has a quantization number “QNO”, four 14 bytes-luminance blocks Y0 to Y3 respectively, and two 10 bytes-chrominance blocks Cr and Cb respectively. Hence, each abovementioned block includes a set of fixed length information such as the DC coefficients and the 8 by 8 block of DCT coefficients through performing the DCT transform of the AC coefficients of the video image. It's noted that the number of bits required to represent the DCT coefficients of each block should be different for respective quantization coding mode. Accordingly, for the fixed size allocation policy, each block will fall into one of following situations: space redundancy and lack.
With reference to FIG. 1, a schematic diagram showing a conventional DV system 10 which receives a DV format data stream from the DV data stream source such as a DTV (digital television), an interactive STB (set-top box), a DV camera and other device that produces the DV stream. The DV format data stream is divided and transmitted as DIF (digital interface format) sequences which is composed of an integral number of fixed length DIF blocks having an identification header for each DIF block that specifies the type of the DIF block and its position in the DIF sequence. Moreover, the DV system 10 includes a DIF Parser 110 for parsing and separating the DIF data into a DV video data stream which is passed to a video decoder 120 for decoding and a DV audio data stream which is passed to an audio decoder 130 for decoding. Later, the decoded video and audio may be represented to the user by displaying on a display and playing on a speaker respectively. In addition, the video decoder 120 further includes a processor 140 for performing the procedure of pre-parse, inverse quantization, and inverses DCT, and a VLD (variable length decoder) 150.
Each frame is divided into several data blocks and then grouped as the “macro-blocks”. During the DV encoding stage, each block is encoded using either an 8 by 8 DCT or 2 by 4 by 8 DCT to produce DCT coefficients of per data block which will be weighted later. Herein, each data block of DCT coefficients is classified into one of four classes with the class number ranged from zero to three, and each coefficient of the data block of DCT coefficients is assigned an area number ranged from zero to three. Accordingly, the FIG. 2A shows the assignment of area numbers to each coefficient in a DCT coefficient matrix while an 8 by 8 DCT is used, and the FIG. 2B shows the assignment while a 2 by 4 by 8 DCT is implemented. In other words, the DCT coefficients are subjected to the quantization procedure for influencing the amount of compressed data accessing and transmitting. The division of each DCT coefficients is determined on the class number, area number and quantization number during quantization stage. It's noted that the quantization number limited the amount of data in one video segment to five compressed macro-blocks in the DV format. Besides, the data block information is passed to the VLD unit during DV decoding so that each data block is decoded to obtain the DCT coefficient which is then passed to the inverse quantization unit and the inverse DCT unit and outputted to the frame storage area for displaying. Herein, the decoding of data block with variable length during a time unit will be inefficient since the digital video is passed to the DV decoder in a constant bit rate.